A mine pollution state monitoring system

Abstract

This invention relates to the field of pollution monitoring technology, specifically a mine pollution status monitoring system. The system includes a concentration mutation identification module, a disturbance frequency calculation module, a pollution area location module, a diffusion path judgment module, and a monitoring result output module. In this invention, by identifying time points with significant differences in concentration time series and marking the location of mutations, the system achieves precise location of sudden pollution events. It achieves time-focused analysis of disturbance behavior by dividing the operation cycle and statistically analyzing the frequency of concentration trend reversals. Combined with the trend consistency of adjacent sensing points, it spatially determines the concentrated pollution area in the mine and determines the continuity of the pollution path based on the continuity of the trend direction. Thus, it dynamically reconstructs the pollution evolution process in both time series and spatial distribution dimensions, and completely marks the pollution source location, propagation link, and affected area in a unified map, achieving the effects of clear mine pollution status location, traceable diffusion path, and intuitive intervention decisions.

Description

Technical Field

[0001] This invention relates to the field of pollution monitoring technology, and in particular to a mine pollution status monitoring system. Background Technology

[0002] The field of pollution monitoring technology involves the identification, detection, and concentration assessment of pollutants in the environment. Its core aspects include pollution source identification, pollutant type analysis, pollution diffusion trend monitoring, and data acquisition and processing. It covers multiple areas such as gaseous pollution detection, particulate matter monitoring, water pollution analysis, and comprehensive environmental indicator assessment, and is widely used in industrial production, urban environmental governance, and mine safety management. Among these, a mine pollution status monitoring system refers to a system used for real-time monitoring of the status of harmful pollutants in underground mine environments. It targets the monitoring and identification of concentrations of harmful gases such as dust, methane, and carbon monoxide within the mine's working space. Typically, it uses fixed electrochemical sensors, light scattering detectors, and catalytic combustion probes to collect pollutant concentration data. Environmental pollution data is acquired through sensors deployed in the mine's main ventilation ducts or work sites and transmitted to a ground-based monitoring platform to achieve the acquisition and updating of pollution status information.

[0003] While existing mine pollution monitoring systems can collect and transmit sensor data in real time, they lack a detailed analysis mechanism for pollution concentration change patterns. This makes it impossible to effectively identify high-risk periods when pollution conditions change abruptly or fluctuate frequently. When the pollution evolution process reflected by the data exhibits complex fluctuations, relying solely on fixed-point concentration values ​​is insufficient to reflect whether the pollution source is persistent or diffuse. This leads to unclear pollution area identification and delayed response. For example, in areas with frequent abnormal gas fluctuations, the inability to identify trend convergence phenomena results in the concentrated distribution of pollution being overlooked, affecting the overall efficiency and accuracy of safety monitoring. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a mine pollution status monitoring system.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a mine pollution status monitoring system comprising:

[0006] The concentration mutation identification module acquires continuous concentration time series data of the underground gas sensor within the sampling period, calculates the gas concentration change value and compares the change difference value. When the change difference value is greater than or equal to the concentration mutation rate threshold, the current time point is marked as the mutation point, and a pollution mutation manifestation location table is generated.

[0007] Based on the pollution mutation manifestation location table, the disturbance frequency calculation module divides the operation cycle into equal-length time periods, detects and counts the number of reversals in the concentration change direction within each period, filters time periods that are greater than or equal to the disturbance frequency threshold, and generates a list of high-frequency disturbance sections.

[0008] The pollution area location module obtains the concentration change trend sequence of the gas sensor at the corresponding spatial location in the corresponding time period according to the list of high-frequency disturbance sections, counts the number of sensors with consistent trends, selects the spatial location that meets the regional concentration judgment benchmark value as the suspected pollution area, and generates a list of suspected pollution areas.

[0009] Based on the list of suspected pollution areas, the diffusion path determination module obtains the concentration trend change direction of gas sensors in the suspected pollution areas for three consecutive time periods before and after the pollution manifestation point. If the trend is continuous and consistent, the corresponding path is marked as a pollution diffusion channel, and a pollution diffusion channel structure diagram is generated.

[0010] Based on the pollution diffusion channel structure diagram, the monitoring result output module draws a distribution map according to the spatial location of the mine, marks all pollution propagation links, and obtains a monitoring map of the mine pollution status.

[0011] As a further aspect of the present invention, the pollution mutation manifestation location table includes the spatial location of the mutation point, the time information of the mutation point, and the mutation intensity level; the disturbance high-frequency segment list includes the high-frequency time period number, the frequent reversal sensor number, and the reversal frequency index; the pollution suspected area list includes the pollution concentration area number, the number of trend consistency sensors in the area, and the suspected pollution level; the pollution diffusion channel structure diagram includes the diffusion path number, the path start and end positions, and the continuous trend maintenance period number; and the mine pollution status monitoring map includes the pollution source manifestation location, the pollution propagation link path map, and the pollution impact area distribution record.

[0012] As a further aspect of the present invention, the concentration mutation recognition module includes:

[0013] The data stream receiving submodule acquires continuous concentration time series data from the downhole fixed-point gas sensor within the sampling period, numbers and stores the concentration data in chronological order to form a structured time series concentration sequence, establishes a data receiving buffer list, and generates a continuous concentration sequence value set.

[0014] The rate of change calculation submodule, based on the continuous concentration sequence value set, selects any three time points to form two adjacent time periods, calculates the gas concentration change value in the two time periods respectively, calculates the difference between the two change values, and combines the current sampling time interval to calculate the concentration change difference value per unit time, generating a concentration change difference rate value set.

[0015] The mutation point determination submodule compares the concentration change difference rate values ​​at each time point with the concentration mutation rate threshold based on the concentration change difference rate value set. When the rate value at a time point is greater than or equal to the concentration mutation rate threshold, the corresponding time point is marked as a mutation point, the location number and time information of the mutation point are recorded, and a pollution mutation manifestation location table is generated.

[0016] As a further aspect of the present invention, the disturbance frequency calculation module includes:

[0017] Based on the pollution mutation manifestation location table, the concentration data extraction submodule locates the sensor number and mutation occurrence time corresponding to each mutation point, extracts the complete gas concentration data sequence of each sensor in the corresponding working cycle, organizes the concentration data in chronological order and unifies the sampling time interval, divides the working cycle into equal-length time periods, and generates a segmented concentration data sequence for the working cycle.

[0018] The reversal count detection submodule traverses the continuous concentration data within each time period according to the segmented concentration data sequence of the operation cycle, compares the concentration change direction of two adjacent time points in turn, records the direction sequence of positive and negative changes, counts the number of consecutive reversals of direction and assigns a corresponding disturbance frequency value to each time period, and generates a segmented disturbance frequency value set.

[0019] The high-frequency segment filtering submodule compares the disturbance frequency value corresponding to each time period with the set disturbance frequency threshold according to the segmented disturbance frequency value set, filters the time periods with disturbance frequency values ​​greater than or equal to the disturbance frequency threshold, extracts the corresponding time period number and the corresponding sensor number information, and generates a list of disturbance high-frequency segments.

[0020] As a further aspect of the present invention, the contaminated area location module includes:

[0021] The concentration trend extraction submodule extracts the concentration data of all sensors at the corresponding spatial locations within each time period based on the list of high-frequency disturbance segments. It constructs a concentration change sequence for each sensor in each time period according to the sampling order, calculates the concentration difference between adjacent time points in each sequence and extracts the difference direction label, establishes the concentration change trend structure of each sensor in each time period, and generates a spatial concentration trend sequence matrix.

[0022] The trend consistency determination submodule compares the direction of the concentration trend vectors corresponding to adjacent sensors in each time period based on the spatial concentration trend sequence matrix. If the trend symbol sequence positions of any two groups of adjacent sensors are consistent, it is determined that the trends are similar. The number of sensors that meet the trend consistency determination conditions is counted in turn, and the cumulative number under the current spatial location unit is recorded to generate a set of spatially consistent sensor count values.

[0023] The spatial clustering identification submodule determines, point by point, whether the number of consistent sensors at each spatial location is greater than or equal to the regional clustering judgment benchmark value based on the set of consistent sensor counts at the spatial location. If the condition is met, the corresponding location is marked as a suspected pollution area, the corresponding spatial location number and the corresponding time period are recorded, and a list of suspected pollution areas is established.

[0024] As a further aspect of the present invention, the diffusion path determination module includes:

[0025] The trend direction extraction submodule extracts the sensor number and the corresponding time period of the pollution manifestation point under each suspected area based on the list of suspected pollution areas. It traces back two time periods and extends forward one time period to construct a concentration trend data set containing three consecutive time periods before and after the pollution manifestation point. It assigns values ​​to the concentration trend direction of each segment in the set and organizes them into a direction sequence matrix to generate a three-segment concentration trend direction sequence matrix.

[0026] The continuous consistency determination submodule performs positional consistency judgment on each group of trend direction sequences based on the three-segment concentration trend direction sequence matrix. If the direction values ​​of the three time periods are consistent, they are marked as consistent paths. The starting and ending index positions of each trend consistent path in spatial location are calculated, and the trend consistency distance is obtained. The path segment index is established with the sensor pairs that are less than the set difference threshold to obtain the trend consistent path segment index set.

[0027] The path structure construction submodule connects the sensor numbers one by one according to the index set of trend-consistent path segments and records each pair of start and end points as a channel path. It then numbers all paths and establishes a start-end point path mapping table to generate a pollution diffusion channel structure diagram.

[0028] As a further aspect of the present invention, the formula for calculating the trend consistency distance is:

[0029] ;

[0030] in, Indicates sensor In the The trend direction value for each time period is assigned a value of +1 or -1. , For sensors Spatial location coordinates, This indicates the distance between trends.

[0031] As a further aspect of the present invention, the monitoring result output module includes:

[0032] Based on the pollution diffusion channel structure diagram, the pollution node extraction submodule extracts the spatial location index and time label of the starting pollution source manifestation point, the path continuity confirmation point, and the pollution impact area at the end of each channel, establishes three types of key node sets, and generates a pollution channel key node index set.

[0033] The spatial path drawing submodule connects each pollution path according to the key node index set of the pollution channel, in the order of node spatial coordinates and time. It connects the starting pollution source manifestation point to the path continuity confirmation point, and then to the end pollution impact area. Each path is drawn with a unique number, and the path segment type is marked with different colors. The node coordinates, connection order and path number index information are recorded to establish a pollution propagation path topology diagram.

[0034] The state map generation submodule overlays all path structures onto the three-dimensional spatial structure map of the mine based on the pollution propagation path topology map, maps path segments to the mine plan view according to spatial layering, marks pollution node attributes, path numbers and flow arrows, draws pollution link distribution map, and establishes a mine pollution state monitoring map.

[0035] As a further aspect of the present invention, the initial pollution source manifestation point is specifically the node in each path where the pollution concentration changes abruptly the earliest; the path continuity confirmation point is specifically the intermediate node in the path where the concentration trend direction is continuous and consistent; and the endpoint pollution impact area is specifically the concentration increase area at the end of the path.

[0036] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0037] In this invention, by identifying time points with significant differences in concentration over time and marking the location of sudden changes, the invention achieves precise identification of sudden pollution events. By dividing the operation cycle and statistically analyzing the frequency of concentration trend reversals, the invention achieves time-focused analysis of disturbance behavior. By combining the trend consistency of adjacent sensing points, the invention spatially determines the concentrated pollution area in the mine and determines the connectivity of the pollution path based on the continuity of the trend direction. Thus, the invention dynamically reconstructs the pollution evolution process in both time series and spatial distribution dimensions, and fully marks the location of pollution sources, propagation links, and affected areas in a unified map. This achieves the effects of clear mine pollution status positioning, traceable diffusion paths, and intuitive intervention decisions. Attached Figure Description

[0038] Figure 1 This is a system flowchart of the present invention;

[0039] Figure 2 This is a flowchart of the concentration mutation recognition module of the present invention;

[0040] Figure 3 This is a flowchart of the disturbance frequency calculation module of the present invention;

[0041] Figure 4 This is a flowchart of the contaminated area location module of the present invention;

[0042] Figure 5 This is a flowchart of the diffusion path determination module of the present invention;

[0043] Figure 6 This is a flowchart of the monitoring result output module of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0045] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0046] Please see Figure 1 A mine pollution status monitoring system includes:

[0047] The concentration mutation identification module acquires continuous concentration time series data from the underground fixed-point gas sensor within the sampling period, selects any three points to form two time periods, calculates the gas concentration change value for each time period, compares the change difference value and makes a judgment based on the concentration mutation rate threshold. When the change difference value is greater than or equal to the concentration mutation rate threshold, the current time point is marked as a mutation point, the location and time information of the mutation point are recorded, and a pollution mutation manifestation location table is generated.

[0048] The disturbance frequency calculation module obtains the complete gas concentration data sequence within the corresponding work cycle based on the pollution mutation manifestation location table, divides the work cycle into equal-length time periods, detects whether the concentration change direction within each period is continuously reversed, counts the number of direction reversals, filters time periods that are greater than or equal to the disturbance frequency threshold, records the time period number and sensor number, and generates a list of high-frequency disturbance sections.

[0049] The pollution area location module obtains the concentration change trend sequence of all sensors at the corresponding spatial location in the corresponding time period based on the list of high-frequency disturbance segments, determines whether there is a trend convergence phenomenon between adjacent sensors in the same time period, counts the number of sensors with consistent trends, selects spatial locations that meet the criteria of being greater than or equal to the regional concentration judgment benchmark value as suspected pollution areas, and generates a list of suspected pollution areas.

[0050] The diffusion path determination module is based on the list of suspected pollution areas. It obtains the concentration trend change direction of all sensors in each suspected pollution area for three consecutive time periods before and after the pollution manifestation point, determines whether the trend direction is continuous and consistent. If it is continuous and consistent, the corresponding path is marked as a pollution diffusion channel, and the start and end positions are recorded to generate a pollution diffusion channel structure diagram.

[0051] The monitoring results output module is based on the pollution diffusion channel structure diagram. It determines the starting pollution source manifestation point, path continuity confirmation point and the endpoint pollution impact area on each channel, draws a distribution map according to the spatial location of the mine, marks all pollution propagation links, and obtains the mine pollution status monitoring map.

[0052] The pollution mutation location table includes the spatial location of the mutation point, the time information of the mutation point, and the mutation intensity level. The list of high-frequency disturbance segments includes the high-frequency time period number, the frequent reversal sensor number, and the reversal frequency index. The list of suspected pollution areas includes the pollution concentration area number, the number of trend consistency sensors in the area, and the suspected pollution level. The pollution diffusion channel structure diagram includes the diffusion path number, the path start and end location, and the continuous trend maintenance period number. The mine pollution status monitoring map includes the pollution source location, the pollution propagation link path map, and the pollution impact area distribution record.

[0053] Please see Figure 2 The concentration mutation recognition module includes:

[0054] The data stream receiving submodule acquires continuous concentration time series data from the downhole fixed-point gas sensor within the sampling period, numbers and stores the concentration data in chronological order to form a structured time series concentration sequence, establishes a data receiving buffer list, and generates a continuous concentration sequence value set.

[0055] To obtain continuous concentration time-series data from a fixed-point underground gas sensor within a sampling period, gas concentration signals need to be collected in real time from the sensor installed underground. The sensor terminal then samples the data at fixed intervals, for example, every 10 seconds. All collected gas concentration values ​​are written into a concentration sequence array in chronological order. Assuming the current monitoring period is 5 minutes, a total of 30 sets of data can be obtained, denoted as . ,in Indicates the first The methane concentration value corresponding to each second, expressed as a volume fraction (%). After collection, the data is uploaded to a ground server via wired or wireless communication, and then categorized according to time stamps. Perform one-to-one mapping and form a data structure For example, the concentration value obtained at a fixed point within 5 minutes is: The data is temporarily stored in the memory structure through a cache list for easy subsequent operation. To address the data compatibility issues between different sensing devices and computing modules, a unified format structure is required to standardize the concentration data and set flags to distinguish the data sources from multiple locations, ultimately generating a continuous concentration sequence value set.

[0056] The rate of change calculation submodule is based on a continuous concentration sequence value set. It selects any three time points to form two adjacent time periods, calculates the gas concentration change value in the two time periods respectively, calculates the difference between the two change values, and combines the current sampling time interval to calculate the concentration change difference value per unit time, generating a concentration change difference rate value set.

[0057] Based on the established continuous concentration sequence value set, select any three adjacent time points. Corresponding gas concentration value Comprising two time periods , Calculate the concentration changes over the two time periods respectively. , This allows us to obtain the rate of change over two time periods. , Calculate its difference rate value For example, in a set of data:

[0058] , , ;

[0059] , ;

[0060] , ;

[0061] , ;

[0062] ;

[0063] The above process proceeds in a sliding window manner throughout the concentration sequence, forming differences in the rate of change across multiple time periods, thus creating an array. Each item represents the degree of difference in the rate of concentration change within adjacent time periods. In order to unify the dimensions of the rate of difference and the judgment criteria, the absolute value processing method is used to exclude the factor of the direction of change, and the time interval is fixed at 10 seconds to finally generate a set of concentration change rate of difference values.

[0064] The mutation point determination submodule compares the concentration change difference rate value set with the concentration mutation rate threshold at each time point. When the rate value at a time point is greater than or equal to the concentration mutation rate threshold, the corresponding time point is marked as a mutation point. The location number and time information of the mutation point are recorded, and a pollution mutation manifestation location table is generated.

[0065] Based on the acquired set of concentration change rate values, for each item... Compared with the preset concentration mutation rate threshold Perform a step-by-step comparison and judgment; if a certain time point satisfies... If a mutation occurs, the point is marked as a mutation point, and its position number in the original data sequence and its corresponding time information are recorded. For example, setting a threshold If a certain point If the threshold is met, the judgment condition is satisfied. This threshold can be set based on the frequency distribution of mutation rates in historical monitoring data. For example, if the rate difference is extracted from 100 historical mutation events, the 95th percentile in the statistical distribution can be used. That is, if 95% of the normal fluctuation rates are less than 0.0028, then a threshold of 0.003 is representative. The actual sampling situation is shown below:

[0066] Table 1. Statistical table of concentration change rate differences

[0067] Time number Current time / s Concentration difference rate (% / s) Is it a mutation point? 8 80 0.002 no 9 90 0.0031 yes 10 100 0.0029 no 11 110 0.0032 yes

[0068] As shown in Table 1, the rate difference between the 9th and 11th seconds both exceeded the set threshold, and therefore were recorded as mutation points; the output structure of the mutation points is a record array. Each item includes the mutation point time, concentration value, and serial number information, and finally a pollution mutation manifestation location table is established.

[0069] Please see Figure 3 The disturbance frequency calculation module includes:

[0070] The concentration data extraction submodule is based on the pollution mutation manifestation location table. It locates the sensor number and mutation occurrence time corresponding to each mutation point, extracts the complete gas concentration data sequence of each sensor in the corresponding operation cycle, organizes the concentration data in chronological order and unifies the sampling time interval, divides the operation cycle into equal-length time periods, and generates segmented concentration data sequences for the operation cycle.

[0071] Based on the pollution mutation location table, the timestamps and corresponding sensor numbers of the mutation points listed in the table are first extracted. By comparing the occurrence time of each mutation point with the work log records, the start and end time range of the work cycle in which the mutation point is located is determined. Then, the original full gas concentration records of the corresponding sensors within this time range are retrieved and organized into a continuous time series according to the second-level sampling period. Subsequently, the entire work cycle is divided into several equally spaced sub-segments according to the set time length. During the division process, the work cycle start time and total duration need to be evenly divided. For example, a 300-second work cycle is divided into 60-second segments, resulting in 5 sub-segments, which are numbered D1 to D5. Then, the concentration sequence within each sub-segment needs to be aligned with the sampling time. If some time points have missing data due to transmission delay or equipment failure, linear interpolation of adjacent time points is used to complete the data, ensuring the data integrity and equal interval characteristics within each time period. At the same time, the start and end time points, the corresponding sensor number, and the segment number of each segment need to be recorded. The following is an example of actual sampling data:

[0072] Table 2. Segmented Concentration Data for Work Cycle

[0073] Segment number Start time / s End time / s Concentration sequence (%) Sensor number D1 0 60 0.32，0.33，0.35，0.34，0.36，... S001 D2 60 120 0.36，0.37，0.38，0.39，0.38，... S001 D3 120 180 0.38，0.37，0.36，0.34，0.35，... S001

[0074] As shown in Table 2, the concentration data in each segment is in the form of a one-dimensional array with a unified sampling interval and time sequence structure. This data structure will be used for disturbance direction analysis in the next module to finally generate a segmented concentration data sequence for the operation cycle.

[0075] The reversal count detection submodule traverses the continuous concentration data within each time period based on the segmented concentration data sequence of the operation cycle, compares the concentration change direction of two adjacent time points in turn, records the direction sequence of positive and negative changes, counts the number of consecutive reversals of direction and assigns a corresponding disturbance frequency value to each time period, and generates a segmented disturbance frequency value set.

[0076] Based on the concentration data sequence segmented by the work cycle, the concentration sequence within each time period is traversed and processed. First, two consecutive time points are grouped together, and their concentration change direction is calculated. If the latter value is greater than the former value, it is defined as positive; otherwise, it is defined as negative. If the two values ​​are equal, no change in direction is recorded. Then, the continuous direction sequence within the current segment is traversed and judged. If the current direction changes from the previous direction, a reversal is recorded. Finally, the number of direction reversals within the time period is accumulated. Assuming the concentration change direction sequence in segment D1 is: positive, positive, negative, negative, positive, negative, then the number of reversals is 4, which is the perturbation frequency value. To avoid double counting, a flag is set to record the current direction status. The flag is updated and counted only when the direction value changes. The perturbation frequency value needs to be recorded below the corresponding segment number. At the same time, a one-to-one mapping relationship between the segment number and the perturbation frequency value needs to be established. Finally, the following structure is constructed:

[0077] Table 3. Segmented Disturbance Frequency Values

[0078] Segment number Concentration change direction sequence Disturbance frequency (times) D1 Front → Front → Back → Back → Front → Back 4 D2 Positive → Positive → Positive → Positive → Positive 0 D3 negative → negative → negative → positive → negative → positive 3

[0079] As shown in Table 3, segment numbers D1 and D3 record 4 and 3 reversal behaviors respectively. This value will be used for threshold filtering in the next submodule to finally generate a set of segmented perturbation frequency values.

[0080] The high-frequency segment filtering submodule compares the disturbance frequency value corresponding to each time period with the set disturbance frequency threshold according to the segmented disturbance frequency value set, filters the time periods with disturbance frequency values ​​greater than or equal to the disturbance frequency threshold, extracts the corresponding time period number and the sensor number information, and generates a list of disturbance high-frequency segments.

[0081] Based on the segmented disturbance frequency value set, the disturbance frequency value of each time period is extracted sequentially, and a disturbance frequency threshold is set as the screening benchmark. The disturbance frequency value of each segment is compared to see if it is greater than or equal to this threshold. If the condition is met, the time period is marked as a high-frequency disturbance segment, and its segment number and sensor number information are recorded. The setting of the disturbance frequency threshold needs to be determined based on the 90th percentile principle in the data distribution. Assuming that after collecting disturbance frequency values ​​from 1000 time periods, 900 values ​​are found to be less than 3, the disturbance frequency threshold is set to 3, meaning that time periods with a disturbance frequency greater than or equal to 3 are high-frequency segments. The screening process uses a sequential comparison method to judge each disturbance frequency value individually and then marks and summarizes them. The results are presented in tabular form as follows:

[0082] Table 4. List of High-Frequency Disturbance Sections

[0083] Segment number Disturbance frequency (times) Is it a high-frequency band? Sensor number D1 4 yes S001 D2 0 no S001 D3 3 yes S001

[0084] Referring to Table 4, the disturbance frequency values ​​for time periods D1 and D3 are 4 and 3 respectively, which meet the screening criteria. Therefore, they are marked as high-frequency disturbance segments, and a list of high-frequency disturbance segments is finally generated.

[0085] Please see Figure 4 The contaminated area location module includes:

[0086] The concentration trend extraction submodule extracts the concentration data of all sensors at the corresponding spatial locations within each time period based on the list of high-frequency disturbance segments. It constructs the concentration change sequence of each sensor in each time period according to the sampling order, calculates the concentration difference between adjacent time points in each sequence and extracts the difference direction label, establishes the concentration change trend structure of each sensor in each time period, and generates a spatial concentration trend sequence matrix.

[0087] Based on the list of high-frequency disturbance bands, the time period number and corresponding sensor spatial location are extracted for each record item in the list. The spatial coordinate index of the sensor in the monitoring system is mapped using the sensor location information. The raw concentration sampling data of each sensor at that spatial location within that time period are obtained and formed into a one-dimensional sequence in chronological order. This sequence is then labeled with positive and negative directions based on the direction of the concentration difference between adjacent points. If the concentration value at a later time is greater than that at a previous time, it is marked as "increasing," and vice versa; if they are equal, it is marked as "stable." This sequence of directional labels is defined as the concentration change trend vector. Subsequently, the trend of all sensors is analyzed. The potential vector is uniformly stored as a two-dimensional structure, where the first dimension is the spatial location number and the second dimension is the trend vector of each sensor at that location. For example, for sensor S101 located in segment A1, its concentration sequence from 120 seconds to 180 seconds is {0.28, 0.31, 0.32, 0.30, 0.29}, and the corresponding trend vector is {increasing, increasing, decreasing, decreasing}, which are represented by numerical codes as {+1, +1, -1, -1}. This trend vector will then serve as the basic data structure for subsequent judgment of trend consistency, and will be uniformly classified according to spatial location to finally construct a structured matrix of concentration trends.

[0088] Table 5 Concentration Trend Sequence Matrix

[0089] Spatial location Sensor number Concentration trend series A1 S101 +1，+1，-1，-1 A1 S102 +1，+1，-1，-1 A2 S201 -1，-1，+1，+1

[0090] As shown in Table 5, sensors at different locations have independent trend sequences. This structure will be input into the next sub-module for consistency determination, generating a spatial concentration trend sequence matrix.

[0091] The trend consistency determination submodule compares the direction of the concentration trend vectors of adjacent sensors in each time period based on the spatial concentration trend sequence matrix. If the trend symbol sequence positions of any two groups of adjacent sensors are consistent, it is determined that the trends are similar. The number of sensors that meet the trend consistency determination conditions is counted in turn, and the cumulative number under the current spatial location unit is recorded to generate a set of spatially consistent sensor count values.

[0092] Based on the spatial concentration trend sequence matrix, pairwise comparisons are performed on the concentration trend sequences of all adjacent sensors within the same spatial location cell. First, any two trend sequences are selected sequentially, and their trend direction label values ​​within the same time period are compared. If a sequence pair has the same direction label at more than half of the time points, the trend sequences are considered consistent. For example, if all positions are completely identical between the sequence pairs {+1, +1, -1, -1} and {+1, +1, -1, -1}, they are considered consistent; if the positions are completely identical between {+1, +1, -1, -1} and {+1, -1}, they are considered consistent. If the sequence {1, 2, 3} is true, then only the 1st and 3rd positions are consistent, which is less than 50% of the four time points, and is therefore considered inconsistent. The number of sensor pairs with consistent trends within the same spatial location is counted. By counting the number of sensor pairs with different consistent trends, the total number of sensors exhibiting consistent trends at that spatial location can be deduced. For example, in location A1, if there are 3 sensors, paired to form 3 trend pairs, and 2 of these pairs meet the consistency judgment condition, then it is determined that there are 3 sensors with consistent trends at that location. The number of consistent sensor values ​​at all locations is recorded, and an index is created between the location number and the corresponding value, forming the following structure:

[0093] Table 6. Statistics of Spatially Consistent Sensors

[0094] Spatial location Number of sensors with consistent trends (units) A1 3 A2 1 A3 4

[0095] Refer to Table 6 to record the trend consistency performance at each spatial location, providing a quantitative basis for subsequent spatial determination, and finally generating a set of sensor quantity values ​​with consistent spatial location.

[0096] The spatial clustering identification submodule determines whether the number of consistent sensors at each spatial location is greater than or equal to the regional clustering judgment benchmark value based on the set of consistent sensor quantity values ​​at spatial locations. If the condition is met, the corresponding location is marked as a suspected pollution area, the corresponding spatial location number and the corresponding time period are recorded, and a list of suspected pollution areas is established.

[0097] Based on the set of sensor counts consistent with spatial location, the statistically consistent sensor counts at each spatial location are read point by point. A regional clustering judgment benchmark is set and compared with the count value at each location. If the value at a spatial location is greater than or equal to the benchmark, the spatial unit is considered to have a pollution trend clustering. This judgment benchmark is derived from historical clustered events, and a commonly used setting value is 3, meaning that at least 3 sensors in the same spatial unit must have consistent trends to be considered a clustering area. During the comparison operation, the spatial locations that meet the conditions are extracted and their corresponding time period identifiers are recorded to form a polluted spatial candidate list. All spatial location numbers that meet the conditions are summarized to construct a pollution location index list. For example, when the count values ​​of locations A1 and A3 are 3 and 4 respectively, both meeting the benchmark requirement, while A2 is only 1, not meeting the benchmark restriction, only A1 and A3 are recorded.

[0098] Table 7. Determination of Suspected Pollution Areas

[0099] Spatial location Consistent trend quantity Judgment benchmark value Is it a suspected area? A1 3 3 yes A2 1 3 no A3 4 3 yes

[0100] As shown in Table 7, A1 and A3 are assigned a suspected label, and a list of suspected contamination areas is finally established.

[0101] Please see Figure 5 The diffusion path determination module includes:

[0102] The trend direction extraction submodule is based on the list of suspected pollution areas. It extracts the sensor number and the time period of the corresponding pollution manifestation point under each suspected area, traces back two time periods and extends forward one time period to construct a concentration trend data set containing three consecutive time periods before and after the pollution manifestation point. It assigns values ​​to the concentration trend direction of each segment in the set and organizes them into a direction sequence matrix to generate a three-segment concentration trend direction sequence matrix.

[0103] Based on the list of suspected contamination areas, the spatial location number and time period index of the pollution manifestation point corresponding to each record are extracted. Taking the pollution manifestation time period as the center, two time periods are traced forward and one time period is extended forward to obtain the index numbers for a total of three time periods. Then, the corresponding sensor number is indexed according to the spatial location, and the concentration sampling sequence of the sensor in each time period is extracted. The concentration sequence is organized according to a fixed sampling frequency; for example, the number of sampling points in each time period is 5, and the sampling frequency is 1Hz. For example, the concentration values ​​of sample data S01 in the t1 to t3 periods are {0.30, 0.31, 0.32, 0.34, 0.35} and {0.3}, respectively. For samples {5, 0.35, 0.36, 0.38, 0.39} and {0.39, 0.38, 0.37, 0.36, 0.35}, the direction is determined based on the difference between adjacent sampling points. If the difference > 0, the direction is +1; if the difference < 0, it is -1; and if they are equal, it is 0. The trend for segment t1 is determined as {+1, +1, +1, +1}. The overall trend direction is the mode of all directions in this segment, so the direction for t1 is +1. Similarly, t2 is +1, and t3 is -1. The final direction sequence for S01 is {+1, +1, -1}. This operation is performed on the sensors in all suspected areas to form a direction vector matrix. Some example data are as follows:

[0104] Table 8. Three-segment concentration trend direction sequence matrix

[0105] Spatial location Sensor number Time period t1 Time period t2 Time period t3 A1 S101 +1 +1 −1 A1 S102 +1 +1 +1 A2 S201 −1 −1 −1 A2 S202 +1 −1 0 A2 S203 −1 −1 −1

[0106] As shown in Table 8, the directional data are uniformly normalized to directional label values ​​through the sampling difference method, and finally form a three-segment concentration trend directional sequence matrix that can be used for comparison and judgment.

[0107] The continuity consistency determination submodule, based on a three-segment concentration trend direction sequence matrix, performs positional consistency judgment on each group of trend direction sequences. If the direction values ​​of the three time periods are consistent, they are marked as consistent paths. The starting and ending index positions of each trend consistent path in spatial location are calculated using the following formula:

[0108] ;

[0109] The trend consistency distance is calculated, and path segment indices are established using sensor pairs with differences less than a set threshold, resulting in a trend-consistent path segment index set; where... Indicates sensor In the The trend direction value for each time period is assigned a value of +1 or -1. , For sensors Spatial location coordinates, This represents the fusion difference formed by the difference in trend direction values ​​plus the Euclidean distance of spatial location, with a threshold set at 2.5. When the value is less than 2.5, the trend direction and spatial continuity are considered to be consistent.

[0110] Based on the three-segment concentration trend direction sequence matrix, a consistency determination operation is performed on the trend sequence of each group of sensors. The consistency condition is defined as the direction values ​​of the three time periods being exactly the same, i.e., {+1, +1, +1} or {−1, −1, −1}. Sensors that meet the condition are extracted, and all possible combination pairs are constructed with their spatially adjacent sensors. Then, the trend difference degree and spatial distance are jointly calculated on the basis of the combination pair. The trend difference value is set as the absolute value of the sum of the differences of the three direction values, and the spatial position difference is set as the two-dimensional Euclidean distance. The two are weighted and combined to form the fusion difference value, which is then substituted into the formula for calculation.

[0111] Taking S101 and S102 as examples, their trend directions are {+1, +1, −1} and {+1, +1, +1} respectively. The difference in trends is:

[0112] ;

[0113] If their spatial coordinates are S01 (x=2.0, y=3.0) and S02 (x=3.5, y=2.0) respectively, then the spatial distance is:

[0114] ;

[0115] Final result:

[0116] ;

[0117] If the preset fusion difference threshold is 2.5, then =3.803>2.5, which does not meet the consistency connection condition, so this combination is discarded; taking S201 and S203 as examples again, the trend direction is the same, both are {−1, −1, −1}, the spatial coordinates are (5.0, 6.0) and (6.0, 6.0) respectively, the trend difference is 0, and the spatial distance is 1.0, then =0 + 1.0 = 1.0 < 2.5, which meets the condition. Therefore, it is recorded as a valid path segment and written into the index structure. Finally, all matching... For combined path segments with conditions <2.5, generate a set of path segment indexes with consistent trends.

[0118] formula The operational logic involves superimposing the trend direction difference term and the spatial location distance term to form a fused difference index Δij, where the trend direction difference term... Used to measure sensors and The difference in concentration change trends across three consecutive time periods is used to quantify the consistency of trend directions by summing the differences in each trend value and taking the absolute value. If the three trends are completely consistent, the value is 0; otherwise, the difference is accumulated to reflect the degree of inconsistency in the trends. The spatial distance term... This is used to measure the spatial distance between sensors in a two-dimensional monitoring area. It is calculated using Euclidean distance to ensure that the physical proximity of adjacent sensor positions is quantified into the fusion index. The two are superimposed by addition. The purpose is to use "trend consistency" and "spatial continuity" as parallel and equally important criteria to jointly determine whether there are connectable diffusion path segments between sensors. Therefore, no weighting factor is set in the formula structure. Instead, the two parameters are reflected by parallel structure and addition of operations. Paths with directional consistency but large spatial span will be discarded due to the increase of Euclidean term, and vice versa. The square root operation is used for the standard definition of Euclidean distance to ensure that the squared spatial difference returns to the original dimension and avoids the imbalance between the direction term and the distance term in scale, so that the entire fusion judgment has symmetry and comparability.

[0119] The trend consistency distance measure quantifies the combined degree of consistency in concentration change trends and spatial proximity between two sensors during pollution monitoring. This measure reflects whether two monitoring points are likely to belong to the same continuous propagation chain in the pollution diffusion path. By definition, this distance measure consists of two parts: first, a directional trend difference term, which measures the degree of difference in the direction of concentration change trends of the two sensors over three consecutive time periods, reflecting the consistency of their behavior during the time evolution process; second, a spatial Euclidean distance term, which measures the actual physical distance between the two sensors in the two-dimensional monitoring area, reflecting their spatial adjacency. By directly superimposing these two values, this distance measure can effectively identify sensor pairs that simultaneously satisfy both trend direction continuity and spatial proximity, thus providing criterion support for the segmented identification of pollution diffusion paths. The smaller the value, the more consistent the trends and the closer the positions, and the more likely it is to form an effective diffusion segment; conversely, a larger value excludes the possibility of connection. Therefore, this indicator is used in the path construction stage to determine whether sensor nodes are connected on the basis of dual continuity of trend and position, and is the core criterion for path segment selection.

[0120] The path structure construction submodule connects sensor numbers segment by segment according to the index order of the trend-consistent path segment index set and records each pair of start and end points as a channel path. It numbers all paths and establishes a start and end point path mapping table to generate a pollution diffusion channel structure diagram.

[0121] Based on the trend-consistent path segment index set, the starting sensor number and ending sensor number recorded in each pair of combinations are extracted one by one, and the corresponding spatial coordinate information is obtained by calling the sensor deployment table. The sensor nodes are combined in order of sensor number to form a path node sequence. Multiple segments are then spliced ​​together in a spatial topology connection method to construct a continuous path linked list structure. If a certain endpoint number appears repeatedly as the starting point in the next path segment, the two segments can be spliced ​​into a longer path. This process is repeated to construct all connectable path channels. During this process, each channel structure is assigned a path number and the starting and ending point numbers are recorded. At the same time, the information is written into the pollution diffusion channel structure diagram index table. Finally, all channel numbers and corresponding path information are summarized to establish a pollution diffusion channel structure diagram.

[0122] Please see Figure 6 The monitoring result output module includes:

[0123] The pollution node extraction submodule, based on the pollution diffusion channel structure diagram, extracts the spatial location index and time label of the starting pollution source manifestation point, path continuity confirmation point, and endpoint pollution impact area in each channel (determining the node with the earliest change in pollution concentration in each path as the starting pollution source manifestation point, extracting the intermediate nodes with a continuous and consistent concentration trend direction in the path as the path continuity confirmation point, and extracting the concentration increase area at the end of the path as the endpoint pollution impact area), establishes three types of key node sets, and generates a pollution channel key node index set;

[0124] Based on the pollution diffusion channel structure diagram, the path number and its corresponding node sequence are read one by one. The spatial index and corresponding time period number of each node on each path are extracted. The pollution manifestation time record table and concentration trend structure table are called in sequence to find the sensor number of the first sensor in the path node where a concentration change occurs and record its location as the starting pollution source manifestation point. Then, the concentration trend sequence of the nodes in the middle segment of the path is called, and the continuity is judged according to the sign of the change in direction between segments. If the trend direction does not change for 3 consecutive segments (e.g., all are +1 or -1), the current node is marked as the path continuity confirmation point. Then, the concentration sequence corresponding to the end node of the path is found. If the concentration value shows a monotonically increasing trend in the last 3 time points (e.g., 0.34→0.36→0.38), the end node is recorded as the end point of the pollution-affected area. The spatial index, path number, and direction label of the above three types of nodes are recorded and summarized into a pollution channel node index structure, as shown in the example below:

[0125] Table 9 Index Set of Key Nodes in Pollution Channels

[0126] Path number Node type Node number Spatial location Concentration trend direction Display time / s P001 Starting point of appearance S21 A3 +1 360 P001 Continuity confirmation point S24 A5 +1 420 P001 End point impact area S27 A9 +1 480

[0127] As shown in Table 9, the three key points of path P001 have been clearly marked as three types of functional nodes, and the final key node index set of the pollution channel is generated.

[0128] The spatial path drawing submodule connects each pollution path according to the index set of key nodes of the pollution channel, in accordance with the spatial coordinates and time sequence of the nodes. It connects the starting pollution source display point to the path continuity confirmation point, and then to the end pollution impact area. Each path is drawn separately with a unique number, and the path segment type is marked with different colors. The node coordinates, connection order and path number index information are recorded to establish a topological structure diagram of the pollution propagation path.

[0129] Based on the key node index set of the pollution pathway, and using the path number as the primary index, the spatial coordinates of three key nodes in each path are extracted. A line is drawn between the starting point and the continuity confirmation point, and the spatial coordinate pairs are recorded as path segment A. Subsequently, the same line drawing process is performed between the continuity confirmation point and the endpoint pollution impact area, resulting in path segment B. Each path constitutes two continuous line segments, stored hierarchically according to path number. Each line segment is assigned a direction label, and the path direction arrows are uniformly marked according to the trend direction. Path segment types are identified by different colors; for example, red represents the source segment, and orange represents the diffusion segment. To facilitate nested recording of the spatial coordinate pairs and path number mapping relationship of each line segment (e.g., S21-S24 is segment A, S24-S27 is segment B), the above structuring process is performed on all paths, ultimately constructing the following path topology table:

[0130] Table 10 Topological Structure of Pollution Transmission Pathways

[0131] Path number Path segment start point End point of path segment Starting point coordinates (m) End point coordinates (m) Path segment type direction P001 S21 S24 （12，30，5） （20，36，5） Source section → P001 S24 S27 （20，36，5） （28，40，5） diffusion segment →

[0132] As shown in Table 10, the complete path of path number P001 has been visualized and structured, and a topological structure diagram of the pollution transmission path has been finally established.

[0133] The state map generation submodule overlays all path structures onto the three-dimensional spatial structure map of the mine based on the topology map of the pollution propagation path, maps the path segments to the mine plan view according to spatial layering, marks the pollution node attributes, path numbers and flow arrows, draws the pollution link distribution map, and establishes a mine pollution status monitoring map.

[0134] Based on the topological structure of the pollution propagation path, the starting and ending coordinates of each path segment are read one by one to obtain the corresponding layered data in the mine spatial structure map. The path segments are then projected onto the main mine roadway plan in three-dimensional spatial coordinates, and all path segments are overlaid onto a unified layer using path numbers as labels. Different graphic symbols are used to label the starting point, ending point, and continuity nodes of each path segment. For example, the source point is labeled with a red dot, the confirmation point with a blue rhombus, and the ending point with a green square. Directional arrows are represented by a unified vector, and the flow direction corresponds to the trend direction. The path number is labeled in black in the middle of the path segment. A legend is added at the bottom of the image to explain the meaning of each graphic. The final overall output view is as follows:

[0135] Table 11 Summary Table of Nodes in Mine Pollution Status Map

[0136] Element Number Node number Spatial coordinates (m) Type Name Path number G001 S21 （12，30，5） Starting point of appearance P001 G002 S24 （20，36，5） Continuity confirmation point P001 G003 S27 （28，40，5） End point impact area P001

[0137] As shown in Table 11, the node attributes corresponding to each graphic element have been sorted out, and the mine pollution status monitoring map has been finally established.

[0138] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A mine pollution status monitoring system, characterized in that the system... include: The concentration mutation identification module acquires continuous concentration time series data of the underground gas sensor within the sampling period, calculates the gas concentration change value and compares the change difference value. When the change difference value is greater than or equal to the concentration mutation rate threshold, the current time point is marked as the mutation point, and a pollution mutation manifestation location table is generated. Based on the pollution mutation manifestation location table, the disturbance frequency calculation module divides the operation cycle into equal-length time periods, detects and counts the number of reversals in the concentration change direction within each period, filters time periods that are greater than or equal to the disturbance frequency threshold, and generates a list of high-frequency disturbance sections. The pollution area location module obtains the concentration change trend sequence of the gas sensor at the corresponding spatial location in the corresponding time period according to the list of high-frequency disturbance sections, counts the number of sensors with consistent trends, selects the spatial location that meets the regional concentration judgment benchmark value as the suspected pollution area, and generates a list of suspected pollution areas. Based on the list of suspected pollution areas, the diffusion path determination module obtains the concentration trend change direction of gas sensors in the suspected pollution areas for three consecutive time periods before and after the pollution manifestation point. If the trend is continuous and consistent, the corresponding path is marked as a pollution diffusion channel, and a pollution diffusion channel structure diagram is generated. Based on the pollution diffusion channel structure diagram, the monitoring result output module draws a distribution map according to the spatial location of the mine, marks all pollution propagation links, and obtains a monitoring map of the mine pollution status. The diffusion path determination module includes: The trend direction extraction submodule extracts the sensor number and the corresponding time period of the pollution manifestation point under each suspected area based on the list of suspected pollution areas. It traces back two time periods and extends forward one time period to construct a concentration trend data set containing three consecutive time periods before and after the pollution manifestation point. It assigns values ​​to the concentration trend direction of each segment in the set and organizes them into a direction sequence matrix to generate a three-segment concentration trend direction sequence matrix. The continuous consistency determination submodule performs positional consistency judgment on each group of trend direction sequences based on the three-segment concentration trend direction sequence matrix. If the direction values ​​of the three time periods are consistent, they are marked as consistent paths. The starting and ending index positions of each trend consistent path in spatial location are calculated, and the trend consistency distance is obtained. The path segment index is established with the sensor pairs that are less than the set difference threshold to obtain the trend consistent path segment index set. The path structure construction submodule connects the sensor numbers one by one according to the index of the trend-consistent path segment index set, records each pair of start and end positions as a channel path, numbers all paths and establishes a start and end path mapping table to generate a pollution diffusion channel structure diagram. The formula for calculating the trend consistency distance is: ； in, Indicates sensor In the The trend direction value for each time period is assigned a value of +1 or -1. , For sensors Spatial location coordinates, This indicates the distance between trends.

2. The mine pollution status monitoring system according to claim 1, characterized in that, The pollution mutation location table includes the spatial location of the mutation point, the time information of the mutation point, and the mutation intensity level. The list of high-frequency disturbance segments includes the high-frequency time period number, the frequent reversal sensor number, and the reversal frequency index. The list of suspected pollution areas includes the pollution concentration area number, the number of trend consistency sensors in the area, and the suspected pollution level. The pollution diffusion channel structure diagram includes the diffusion path number, the path start and end positions, and the continuous trend maintenance period number. The mine pollution status monitoring map includes the pollution source location, the pollution propagation link path map, and the pollution impact area distribution record.

3. The mine pollution status monitoring system according to claim 1, characterized in that, The concentration mutation identification module includes: The data stream receiving submodule acquires continuous concentration time series data from the downhole fixed-point gas sensor within the sampling period, numbers and stores the concentration data in chronological order to form a structured time series concentration sequence, establishes a data receiving buffer list, and generates a continuous concentration sequence value set. The rate of change calculation submodule, based on the continuous concentration sequence value set, selects any three time points to form two adjacent time periods, calculates the gas concentration change value in the two time periods respectively, calculates the difference between the two change values, and combines the current sampling time interval to calculate the concentration change difference value per unit time, generating a concentration change difference rate value set. The mutation point determination submodule compares the concentration change difference rate values ​​at each time point with the concentration mutation rate threshold based on the concentration change difference rate value set. When the rate value at a time point is greater than or equal to the concentration mutation rate threshold, the corresponding time point is marked as a mutation point, the location number and time information of the mutation point are recorded, and a pollution mutation manifestation location table is generated.

4. The mine pollution status monitoring system according to claim 1, characterized in that, The disturbance frequency calculation module includes: Based on the pollution mutation manifestation location table, the concentration data extraction submodule locates the sensor number and mutation occurrence time corresponding to each mutation point, extracts the complete gas concentration data sequence of each sensor in the corresponding working cycle, organizes the concentration data in chronological order and unifies the sampling time interval, divides the working cycle into equal-length time periods, and generates a segmented concentration data sequence for the working cycle. The reversal count detection submodule traverses the continuous concentration data within each time period according to the segmented concentration data sequence of the operation cycle, compares the concentration change direction of two adjacent time points in turn, records the direction sequence of positive and negative changes, counts the number of consecutive reversals of direction and assigns a corresponding disturbance frequency value to each time period, and generates a segmented disturbance frequency value set. The high-frequency segment filtering submodule compares the disturbance frequency value corresponding to each time period with the set disturbance frequency threshold according to the segmented disturbance frequency value set, filters the time periods with disturbance frequency values ​​greater than or equal to the disturbance frequency threshold, extracts the corresponding time period number and the corresponding sensor number information, and generates a list of disturbance high-frequency segments.

5. The mine pollution status monitoring system according to claim 1, characterized in that, The contaminated area location module includes: The concentration trend extraction submodule extracts the concentration data of all sensors at the corresponding spatial locations within each time period based on the list of high-frequency disturbance segments. It constructs a concentration change sequence for each sensor in each time period according to the sampling order, calculates the concentration difference between adjacent time points in each sequence and extracts the difference direction label, establishes the concentration change trend structure of each sensor in each time period, and generates a spatial concentration trend sequence matrix. The trend consistency determination submodule compares the direction of the concentration trend vectors corresponding to adjacent sensors in each time period based on the spatial concentration trend sequence matrix. If the trend symbol sequence positions of any two groups of adjacent sensors are consistent, it is determined that the trends are similar. The number of sensors that meet the trend consistency determination conditions is counted in turn, and the cumulative number under the current spatial location unit is recorded to generate a set of spatially consistent sensor count values. The spatial clustering identification submodule determines, point by point, whether the number of consistent sensors at each spatial location is greater than or equal to the regional clustering judgment benchmark value based on the set of consistent sensor counts at the spatial location. If the condition is met, the corresponding location is marked as a suspected pollution area, the corresponding spatial location number and the corresponding time period are recorded, and a list of suspected pollution areas is established.

6. The mine pollution status monitoring system according to claim 1, characterized in that, The monitoring result output module includes: Based on the pollution diffusion channel structure diagram, the pollution node extraction submodule extracts the spatial location index and time label of the starting pollution source manifestation point, the path continuity confirmation point, and the pollution impact area at the end of each channel, establishes three types of key node sets, and generates a pollution channel key node index set. The spatial path drawing submodule connects each pollution path according to the key node index set of the pollution channel, in the order of node spatial coordinates and time. It connects the starting pollution source manifestation point to the path continuity confirmation point, and then to the end pollution impact area. Each path is drawn with a unique number, and the path segment type is marked with different colors. The node coordinates, connection order and path number index information are recorded to establish a pollution propagation path topology diagram. The state map generation submodule overlays all path structures onto the three-dimensional spatial structure map of the mine based on the pollution propagation path topology map, maps path segments to the mine plan view according to spatial layering, marks pollution node attributes, path numbers and flow arrows, draws pollution link distribution map, and establishes a mine pollution state monitoring map.

7. The mine pollution status monitoring system according to claim 6, characterized in that, The initial pollution source manifestation point is specifically the node where the pollution concentration changes earliest in each path; the path continuity confirmation point is specifically the intermediate node in the path where the concentration trend direction is continuous and consistent; and the endpoint pollution impact area is specifically the concentration increase area at the end of the path.

Images